Daniel came out on that Sunday, and he walked the lane with me the way he walks a field for the Soil and Water Conservation District, slowly, looking down, stopping to press his boot into the soil and to look at the way water had settled in the low places. He explained, again, patiently, what I already knew but needed to hear in his professional voice: the tile drains under the swale carried water away only so fast, the topsoil over the basin was deep and fine and held water like a sponge, and once it was saturated, the suction in that mud could hold a wheel tighter than any rut. He said a tractor had gone down there in 1998, and Earl Thompson, who farms the parcel to the north, had told the same story at the kitchen table more than once, about a wet October and a tractor that had to be winched out with two others, and a driver who walked home muddy to his knees and grateful not to be hurt. Earl was seventy-one, and he did not exaggerate. When Earl said ground was bad, ground was bad. Daniel measured the moisture in the swale with a probe that afternoon, and he showed me the reading, and he said the basin was already holding more water than the surrounding field, even before the real autumn rains. I believed him. I had believed him in June. The signs believed him too. They stood in a row along the lane, weathered and plain, and they said what they had always said.
The rain week came in the middle of August, the kind of week Iowa gives you when summer is deciding whether to let go. It did not arrive as one storm. It arrived as a procession: a heavy shower on Monday afternoon that turned the lane to slick mud, a slow rain all Tuesday that soaked in rather than running off, a thunderstorm on Wednesday night that rattled the windows and left standing water in every low place, and then two more days of drizzle that kept everything wet without ever quite washing it clean. The corn loved it, in the way corn loves a drink after a dry spell, and the leaves turned a deeper green and the ears filled out. The swale did not love it. The swale drank and drank and stayed full. Daniel came out again on the Friday of that week, after the drizzle had finally lifted, and he walked to the basin with his probe and his boots, and he did not need the probe to tell him what his boots told him first. His foot sank past the ankle in the soft ground at the edge, and when he pulled it free, the mud made a sound like a mouth closing. He measured anyway, because Daniel measures, and he wrote the number in his notebook and showed it to me, and he said the saturation was as high as he had seen it outside of spring thaw. He said, in plain language, that no vehicle belonged on that ground until it had dried for days, and that even then the basin itself should be avoided, because the grass strip over it, the firm-looking strip that tempted a driver in a hurry, was the worst place of all. It looked like solid ground because the grass held it together on top. Underneath, it was soup.
I listened, and I did what a farmer does when a warning is not enough. I added to the warnings. Margaret and I drove the pickup to the barn and loaded orange survey stakes, the bright kind you cannot miss even at dusk, and a roll of flagging tape, and we drove out to the swale and set the stakes in a line along both sides of the grass strip, close enough together that no honest driver could claim to have missed them. Then I went back for the hay bales. I had a row of round bales left from the first cutting, sitting near the barn, and I rolled six of them, with the tractor and a great deal of slow manoeuvring, to the edge of the swale, and I set them as a barricade across the lane where it narrowed toward the basin. They were not a wall. They were a statement, heavy and round and obvious, the kind of obstacle you have to make a decision to move. A person who drives around a gate can tell herself the gate was left open. A person who moves a hay bale has decided, with both hands, to go through anyway. I wanted that decision to be visible, if it came, and I wanted the trail camera to see it. The camera was already up, mounted on a fence post at the lane's bend, aimed at the swale approach, and Margaret checked its batteries that evening and set it to record on motion. I locked the gate again, I tested the chain, and I went home and wrote it all in the logbook: stakes set, bales placed, camera checked, ground saturated, rain total for the week noted from the gauge by the barn.